The demand for 360° video streaming has been growing at a phenomenal rate due to the emergence of virtual and mixed reality (VR/MR). The interactive nature of VR/MR poses real-time challenges for 360° video transcoding and streaming. In this thesis, we tackle two challenges. Firstly, we proposed a novel distributed transcoding scheme to speed up the transcoding time in a resource-efficient manner, by exploring transcoding parallelization and orchestration of Docker containers. The goal is to push for real-time production of 360° videos in higher quality. Real-time transcoding opens the opportunity for dynamically adjusting the transcoding setting for the best visual quality subject to bandwidth limit and client-side status, instead of a fixed set of quality levels throughout a streaming session. Secondly, we exploited this opportunity and proposed a new adaptive streaming algorithm that directs the transcoding to produce the transcoding segment that best suits the current streaming condition. Our analysis and evaluation show that the new adaptive algorithms lead to higher visual quality by balancing the bandwidth requirements across different regions of a segment during a streaming session in comparison to the conventional adaptive streaming algorithm.